Working Paper Series Dominic Quint, Fabrizio Venditti The influence of OPEC+ on oil prices: a quantitative assessment No 2467 / September 2020 Disclaimer: This paper should not be reported as representing the views of the European Central Bank (ECB). The views expressed are those of the authors and do not necessarily reflect those of the ECB. Abstract Between January 2017 and March 2020 a coalition of oil producers led by OPEC and Russia (known as OPEC+) cut oil production in an attempt to raise the price of crude oil. In March 2020 the corona virus shock led to a collapse of this coalition, as members did not agree on keeping the oil market tight in the face of a large negative demand shock. Yet, was OPEC+ actually effective in sustaining the price of oil? Between 2017 and early 2020 when the OPEC+ strategy was in place, oil inventories fell substantially and the price of oil reached a peak of around 80 USD per barrel, from a minimum of 30 USD in 2016. This suggests that the OPEC+ strategy had a significant impact on the global oil market. Yet, to what extent did crude prices actually reflect OPEC+ production cuts rather than other factors, like swings in demand for oil? How would the price of oil have evolved had OPEC+ not cut supply? This paper provides an answer to these questions through a counterfactual analysis based on two structural models of the global oil market. We find the impact of OPEC+ on the market was overall quite limited, owing to significant deviations from the assigned quotas. On average, without the OPEC+ cuts, the price of oil would have been 6 percent (4 USD) lower. JEL Classification Codes: Q43, C53 Keywords: Oil price, OPEC, oil supply, oil demand, shale oil ECB Working Paper Series No 2467 / September 2020 1 Non-technical summary The advent of US shale oil has substantially changed the structure of global oil production. The emergence of the US as a dominant market player has had profound implications for the strategic behaviour of other oil producers, and in particular for those coalesced in the Organiza- tion of Petroleum Exporting Countries (OPEC). OPEC, which includes 14 members including large producers like Saudi Arabia, Iraq, Iran and the United Arab Emirates, accounted for around 43 percent of global oil production in 2008. In 2019 its share had fallen to 39, as the US increased its oil market share from 8 to 15 percent. Not only the US are by now the largest oil producers, they have also become oil exporters, after a ban that prevented producers to sell oil abroad was lifted in December 2015. In 2016, in an attempt to regain some control over crude prices, OPEC and a number of other non-US oil producers forged an alliance, known as OPEC+. Among non-OPEC members, this coalition included large producers like Russia (the second largest oil producer), Mexico and Kazakhstan. Overall, OPEC+ accounted for around 45 percent of global crude production. The strategy adopted by OPEC+ consisted of setting explicit production targets for each member with the aim of bringing oil inventories down to their 2010-2014 average. In March 2020 the OPEC+ coalition broke down. Follow- ing a general risk-off sentiment related to the spreading of the corona virus, the price of oil collapsed. When OPEC suggested the implementation of further production cuts to lift the price of crude, Russia refused to cooperate arguing that US producers would gain the most from new efforts to prop up prices. The clash between Saudi Arabia and Russia meant the end of the OPEC+ coalition. This paper takes stock of the impact that OPEC+ had on the global oil market while it operated, between the end of 2016 and the beginning of 2020. Us- ing Structural Vector Autoregressions we construct counterfactual scenarios that allow us to quantify how oil production and the price of oil would have evolved, had this agreement not been in place. The empirical strategy consists of computing a counterfactual path of global oil production assuming that all the fall in the production of OPEC+ after December 2016 can be attributed to an exogenous shift in their oil supply. Our counterfactual path of oil production is then constructed assuming that OPEC+ would have kept production steady at the level recorded before OPEC+ started implementing production cuts in line with the agreed upon targets. To evaluate the effects of these production cuts on the price of crude we use two complementary specifications. The former is a small model in which we pool together OPEC+ ECB Working Paper Series No 2467 / September 2020 2 production with that of other producers. In the second specification we split the production of OPEC+ from that of the rest of the world, allowing for strategic interactions between these two large oil producing blocks. We find that the impact of OPEC+ on the price of oil varied over time, together with the cohesion of the coalition, and that it was overall quantitatively modest. Averaging over the whole period under analysis, our results indicate that the price of oil would have been around 4 USD per barrel lower, had OPEC+ not cut production. It would have taken a much deeper cut in oil production and a much stronger cohesion to achieve the ambitious target that the coalition had set for itself. ECB Working Paper Series No 2467 / September 2020 3 1 Introduction The advent of US shale oil has substantially changed the structure of global oil production. The emergence of the US as a dominant market player has had profound implications for the strategic behaviour of other oil producers, and in particular for those coalesced in the Organization of Petroleum Exporting Countries (OPEC). OPEC, which is composed by 14 members including large producers like Saudi Arabia, Iraq, Iran and the United Arab Emirates, accounted for around 43 percent of global oil production in 2008 , see Figure 1 (left panel). In 2019 its share had fallen to 39, as the US had increased its oil market share from 8 to 15 percent. The US is by now not only the largest oil producer, but also an important oil exporter, after a ban that prevented producers to sell oil abroad was lifted in December 2015. As a result, the pricing power of OPEC members weakened substantially, oil prices plunged and a supply glut emerged, as testified by a persistent increase in oil inventories. In 2016, in an attempt to regain some control over crude prices, OPEC and a number of other non-US oil producers forged an alliance, known as OPEC+. Among non-OPEC members, this coalition included large producers like Russia (the second largest oil producer), Mexico and Kazakhstan. Overall, OPEC+ accounted for around 45 percent of global crude production. The strategy adopted by OPEC+ consisted of setting explicit production targets for each member with the aim of bringing oil inventories down to their 2010-2014 average. In March 2020 the OPEC+ coalition broke down. Following a general risk-off sentiment related to the spreading of the corona virus, the price of oil collapsed. When OPEC suggested the implementation of further production cuts to lift the price of crude, Russia refused to cooperate arguing that US producers would gain the most from new efforts to prop up prices. The clash between Saudi Arabia and Russia meant the end of the OPEC+ coalition. Yet was the OPEC+ coalition actually effective in steering the price of oil while it operated? A glance at raw data indicates that, while the coalition was in place, the price of crude rose substantially, suggesting a significant role for the supply cuts implemented by these oil produc- ECB Working Paper Series No 2467 / September 2020 4 ers. Yet inventories remained high compared to their historical average, somewhat confuting the idea that OPEC+ achieved their target of substantial market tightening. And oil demand played a key role in keeping oil prices high as economic activity kept growing at a robust pace over these three years, providing an important confounding factor. In this paper we provide a quantitative evaluation on the actual impact of the OPEC+ alliance on the global oil market between January 2017 and early 2020 by constructing coun- terfactual scenarios that allow us to quantify how oil production and the price of oil would have evolved, had this agreement not been in place. These scenarios are constructed and evaluated using structural econometric models of the oil market. Such models have now a long tradition, since the seminal work by Kilian (2009) and Kilian and Murphy (2012, 2014). In our models we follow broadly Kilian and Murphy (2014) and use sign restrictions in combination with elasticity bounds to identify structural oil price shocks in structural vector autoregressions. Our work is close in spirit to Kilian (2017), who quantifies the effect of shale production on the production of Saudi Arabia and the global oil price between 2008 and 2015. The empirical strategy adopted by Kilian (2017) consists of computing a counterfactual path of global oil production assuming (i) that US production would have remained constant rather than rising after 2008 (ii) that the increase in the output of shale producers between 2008 and 2015 was due entirely to exogenous supply shocks, rather than being a reaction to favourable demand conditions. Given these two assumptions one can easily recover the implied path of oil production and oil prices that would have cleared the market, given an estimated structural model. We adopt the same philosophy and assume that all the fall in the production of OPEC+ producers can be attributed to an exogenous shift in their oil supply.